Pancreatic cancer has a dismal prognosis and conventional therapeutic approaches have had minimal impact to improve the course of this aggressive disease. Clearly, a better understanding of the molecular biology of pancreatic cancer is urgently needed to develop novel therapeutic strategies. The tumor-associated stroma is a hallmark characteristic of pancreatic cancer, yet little is known about its role in tumor progression. The same dense stroma is present in pancreatitis, which is a risk factor for pancreatic cancer. Our overall hypothesis is that a secreted factor produced by the stroma, periostin (PN), is important in pancreatitis and subsequent progression to pre-invasive and invasive pancreatic cancer. To test this hypothesis, we will use primary pancreatic stellate cells isolated from human pancreatic cancer-associated stroma in conjunction with state-of- the art animal models and mechanistic approaches to define the role of PN in the regulation of stromal fibroblasts, pancreatic cancer cells and tumor-stromal interactions. Our specific aims are: (1) To identify the effects of PN on pancreatic tumor progression and clinical outcome, (2) to determine the role of PN in the development of fibrosis in pancreatitis and in the progression from pancreatitis to pancreatic cancer, and (3) to characterize the association between PN and CXCL12. The candidate's immediate goals are to gain expertise in signaling mechanisms important in cancer and also to acquire new skills in mouse models of pancreatitis and pancreatic cancer. With the additional training, the candidate's long term goal is to develop an independent research career as a molecular and cell biologist focused on the role of the stroma in pancreatitis and its progression to pancreatic cancer. The candidate is well-prepared to achieve these goals by working with a mentor who is an internationally recognized expert in pancreatitis and pancreatic cancer. Moreover, he has successfully mentored two other K08 awardees in the past. In addition, the candidate will collaborate with well-established scientists who have substantial expertise in signal transduction and transgenic mouse models. Finally, the institutional environment at the University of Texas-M.D. Anderson Cancer Center provides an ideal setting to assure that the scientific and training goals of this application will be met. PUBLIC HEALTH RELEVANCE: Pancreatic cancer is a highly aggressive malignancy with essentially no effective treatment to date. This application focuses on the cells in the stroma, or tumor microenvironment, which are critical partners in cancer progression and metastasis but their precise mechanisms of action are poorly understood. A better understanding of the tumor-associated stroma may provide novel targets for treatment of pancreatic cancer.